Abstract: A signal compression system includes a coder and a decoder. The coder includes an extract unit for extracting an input feature vector from an input signal, a coder memory unit for storing a predesigned vector quantization (VQ) table for the coder such that the coder memory unit uses a set of primary indices to address entries within the pre-designed VQ table, a coder mapping unit for mapping indices from a set of secondary indices to the first set of indices, and a search unit for searching for one index out of the set of secondary indices, wherein the index from the set of secondary indices corresponds to an entry in the coder memory unit, and the entry best represents the input feature vector according to some predetermined criteria.

Abstract: In a CELP coder a comparison between a target signal and a plurality of synthetic signals is made. The synthetic signal is derived by filtering a plurality of excitation sequences from a one dimensional codebook by a synthesis filter having parameters derived from the target signal. The excitation signal which results in a minimum error between the target signal and the synthetic signal is selected. In order to reduce the complexity of the search for the best excitation signal, the selection is done in two stages. First a preselection of a small number of excitation sequences is made by selecting only every L.sup.th codebook entry for preselecting a plurality of excitation sequences. Thereafter, with this small number of excitation sequences, a full complexity search is made in which all excitation sequences surrounding the preselected ones are involved in the selection.

Abstract: An MPEG audio decoder includes a Vector FIFO buffer and a windowed polyphase filter. Groups of vector samples are zeroed out prior to storage (or after storage, if desired) in the Vector FIFO buffer when error concealment is performed.

Abstract: If the same parameter is repeatedly used in an unvoiced frame inherently devoid of pitch, there is produced a pitch of the frame length period, thus producing an extraneous feeling. This can be prevented from occurring by evading repeated use of excitation vectors having the same waveform shape. To this end, when decoding an encoded speech signal obtained on waveform encoding an encoding-unit-based time-axis speech signal obtained on splitting an input speech signal in terms of a pre-set encoding unit on the time axis, input data is checked by CRC by a CRC and bad frame masking circuit 281, which processes a frame corrupted with an error with bad frame masking of repeatedly using parameters of a directly previous frame. If the error-corrupted frame is unvoiced, an unvoiced speech synthesis unit 220 adds the noise to an excitation vector from a noise codebook or randomly selects the excitation vector of the noise codebook.

Abstract: In the disclosed speech decoding device, activation of a postfilter process is halted during unvoiced sections. However, the updating process of the internal states of the postfilter continues even though the postfilter process is not activated during unvoiced sections. At changes between voiced and unvoiced sections, output signals outputted during voiced sections that have been subjected to a postfilter process and output signals outputted during unvoiced sections that have not been subjected to a postfilter process are interpolated and outputted. In one embodiment, a prefilter controller activates a prefilter state updater for unvoiced sections to update the internal state of the filter of a prefilter section based on excited signals to decrease any perception of noncontinuity in an output signal switching between activation and deactivation of the prefilter when changing between voiced and unvoiced sections.

Abstract: A speech packet transmission system of the present invention has a speech coding device and a speech decoding device. In the coding device a speech coder (21) codes a PCM (Pulse Code Modulation) speech and outputs the resulting speech coded data and prediction coefficients. A speech detector (22) determines whether the input PCM speech is voiced or unvoiced. A transmission prediction coefficient memory (23) memorizes the prediction coefficients. A delay circuit (24) delays the speech coded data by a preselected delay time. On the transition from an unvoiced state to a voiced state, a transmitter (25) sends the prediction coefficients as a leading packet and then sends the delayed speech coded data as the following packet. In the decoding device, a receiver (31) received the packets from the coding device separates the packets into the prediction coefficients and the coded data.

Abstract: A variable dimension vector quantization method that uses a single "universal" codebook. The method can be given the interpretation of sampling full-dimensioned codevectors in the universal codebook and generating subcodevectors of the same dimension as input data subvector, which dimension may vary in time. A subcodevector is selected from the codebook to have minimum distortion between it and the input data subvector. The subcodevector with minimum distortion corresponds to the representative, full-dimensioned codevector in the codebook. The codebook is designed by inverse sampling of training subvectors to obtain full-dimension vectors, then iteratively clustering the training set until a stable centroid vector is obtained.

Abstract: Wide-band speech signals and also music signals are coded with relatively less computational efforts and less sound quality deterioration even at low bit rates. A spectral parameter calculator obtains a spectral parameter from sub-frames of an input signal from a sub-frame divider, and quantizes the obtained spectral parameter. A divider divides the difference result from a subcontractor into a plurality of sub-bands. Adaptive codebook circuits obtain a pitch prediction signal by obtaining pitch data in at least one of the sub-bands. Judging circuits execute pitch prediction judgment by using the pitch data in at least one of the sub-bands. A synthesizer synthesizes a pitch prediction signal. A subtractor subtracts the pitch prediction signal from the difference result obtained from a subtractor and thus obtains an excitation signal. An excitation quantizer quantizes the excitation signal with reference to an excitation codebook.

Abstract: A speech coder, formed with a digital speech encoder and a digital speech decoder, utilizes fast excitation coding to reduce the computation power needed for compressing digital samples of an input speech signal to produce a compressed digital speech datastream that is subsequently decompressed to synthesize digital output speech samples. Much of the fast excitation coding is furnished by an excitation search unit in the encoder. The search unit determines excitation information that defines a non-periodic group of excitation pulses The optimal location of each pulse in the non-periodic pulse group is chosen from a corresponding set of pulse positions stored in the encoder. The search unit ascertains the optimal pulse positions by maximizing the correlation between (a) a target group of filtered versions of digital input speech samples provided to the encoder for compression and (b) a corresponding group of synthesized digital speech samples.

Abstract: An encoding apparatus in which an input speech signal is divided into blocks and encoded in units of blocks. The encoding apparatus includes an encoding unit for performing CELP encoding having a noise codebook memory containing having codebook vectors generated by clipping Gaussian noise and codebook vectors obtained by learning using the code vectors generated by clipping the Gaussian noise as initial values. The encoding apparatus enables optimum encoding for a variety of speech configurations.

Abstract: A method and apparatus for providing high-speed data compressing without sacrificing the quality of data reconstruction. Each input vector or block of original data is expressed as a combination of a codebook index and an error differential, or as a compressed version of the original block of data, depending on whether the total number of bits needed to express the input vector as a combination of a codebook index and an error differential is less than the total bits needed to send the compressed version of the original block of data. In one embodiment, the flexibility to express video data in a compressed format or as a combination of a codebook index plus an error differential is provided through a video system employing a codebook, a scalar quantizer, and an entropy coder.

Abstract: The invention provides a speech coding apparatus by which a good sound quality can be obtained even when the bit rate is low. The speech coding apparatus includes an excitation quantization circuit which quantizes an excitation signal using a plurality of pulses. The position of at least one of the pulses is represented by a number of bits determined in advance, and the amplitude of the pulse is determined in advance depending upon the position of the pulse.

Abstract: An input speech signal is encoded as one or more reflection coefficients. To reduce storage requirements, the reflection coefficients are scalar quantized by storing an N-bit code rather than the entire reflection coefficient. An exemplary value for N is 8. A table is provided having 2.sup.N reflection coefficient values. The N-bit code is used to look up reflection coefficient values from the table. To reduce spectral distortion due to scalar quantization, the reflection coefficient values in the table are non-linearly scaled.

Abstract: A block length judging circuit 120 switches block lengths based on a feature quantity obtained from an input signal. A transform circuit 200 executes transform of the signal into frequency components according to the block length. A masking threshold calculating circuit 250 calculates a masking threshold simulating the masking characteristic of psychoacoustical property for each predetermined intra-block section. An inter-block/intra-block bit assignment circuit 300 executes inter-block bit number assignment and/or intra-block bit number assignment to each predetermined intra-block section. A vector quantization circuit 350 vector quantizes transform signal by switching codebooks 360.sub.1 to 360.sub.N according to the assignment bit number, and also quantizes gain by using a gain codebook 370.

Abstract: A speech signal encoding/decoding method is provided. The method of encoding LPC coefficients includes dividing the nth-order line spectral frequencies into lower, middle and upper code vectors, quantizing the middle code vectors using a middle code book to generate a first index, selecting one of a plurality of lower code books according to the lowermost line spectral frequency of the middle code vector and the line spectral frequencies of the lower code vectors, and quantizing the lower code vectors using the selected lower code book to generate a second index, selecting one of a plurality of upper code books according to the uppermost line spectral frequency of the middle code vector and the line spectral frequencies of the upper code vectors, quantizing the upper code vectors using the selected upper code book to generate a third index, and transmitting the first, second and third indexes.

Abstract: A speech synthesis system in which coefficients of a speech synthesis filter are quantized. An LSP or other filter coefficient representation which evolves slowly with time is generated for each of a series of N input speech frames to produce p coefficients in respect of each frame. The coefficients related to the N frames define a p.times.N matrix, with each row of the matrix containing N coefficients and each coefficient of one row being related to a respective one of the N frames. The matrix is split into a series of submatrices each made up from one or more of the rows, and each submatrix is vector quantized independently of the other submatrices using a composite time/spectral weighting function which for example emphasises distortion associated with high energy regions of the spectrum of each of the N input speech frames and is also proportional to the energy and degree of voicing of each of the N input speech frames.

Abstract: A method and apparatus for encoding an input signal, such as a broad-range speech signal, in which a number of decoding operations with different bit rates are enabled for assuring a high encoding bit rate and for minimizing deterioration of the reproduced sound even with a low bit rate. The signal encoding method includes a band-splitting step for splitting an input signal into a number of bands and a step of encoding signals of the bands in a different manner depending on signal characteristics of the bands. Specifically, a low-range side signal is taken out by a low-pass filter from an input signal entering a terminal, and analyzed for Linear Predictive coding by an Linear Predictive coding analysis quantization unit. After finding the Linear Predictive coding residuals, as short-term prediction residuals by an Linear Predictive coding inverted filter, the pitch is found by a pitch analysis circuit. Then, pitch residuals are found by long-term prediction by a pitch inverted filter.

Abstract: A speech encoding method and apparatus including analyzing, using a codebook expressing speech parameters within a predetermined search range, an input speech signal in an audibility weighting filter corresponding to a pitch period longer than the search range of the codebook, and searching, from the codebook, on the basis of the analysis result, a combination of speech parameters by which the distortion of the input speech signal is minimized, and encoding the combination. The apparatus uses an adaptive codebook of pitch and a noise codebook. The codebooks search a group formed by extracting vectors of predetermined length from one original code vector, while sequentially shifting position so that the vectors overlap each other. The search group is further restricted and another preselection is made before the final search. Search is based on inversely convoluted, orthogonally transformed vectors.

Abstract: An apparatus for improving the link margin of a communication link includes a variable rate vocoder which decreases the output bit stream rate it produces so as to reduce the amount of information having to be transmit in the communication link. In one embodiment, the variable rate vocoder includes a plurality of vocoder portions, each of which produces a different bit stream rate. The selector is used for selecting among the output bit streams produced by each vocoder. In another embodiment, a logic device is coupled to the output of the vocoder. The logic device, upon receipt of a control signal, truncates the less important bits.The method for improving link margin includes reducing the vocoder output rate thereby reducing the amount of data being transmit in an communication link. The method also includes using increased error correction coding and transmitting at increased per bit power levels to increase link margin.

Abstract: Harmonics coefficients are estimated in primary coefficients of an orthogonal transform of a speech or a music input signal by using a pitch frequency extracted from the input signal and are quantized into a harmonics code vector. Residue coefficients are calculated by removing the harmonics coefficients from the primary coefficients and quantized into residue code vectors and gain code vectors. It is possible to search harmonics excitation pulses at the harmonics locations for harmonics quantization into the harmonics code vector. On the other hand, it is possible to estimate the harmonics coefficients or excitation pulses by using quantized LSP parameters and to calculate secondary coefficients for use in weighting the harmonics quantization and residue quantization and, if applicable, in excitation pulse search.

Abstract: An encoding/decoding system employing vector quantization realizes a high quality encoding and decoding with decreased quantizing errors, employing a small sized codebook which faithfully represents each of the inputted waveform vectors. An encoding/decoding system includes an encoding apparatus and a decoding apparatus, each having a codebook for storing information vectors representative of a predetermined number of signal patterns and index that determine the information vectors. The encoding apparatus compares a vector representing an object signal to be quantized with each information vector in the codebook, selects an information vector that is closest to the vector and outputs an index for the information vector. The decoding apparatus obtains an information vector corresponding to the index obtained at the encoding apparatus side by referring to the codebook and decodes the object signal. The codebook utilizes a temporary memory connected thereto.

Abstract: Method and system aspects for linear predictive speech encoding are disclosed. These aspects comprise the definition of an error function, the computation of an optimal vector of continuous pitch coefficients together with an optimal pitch, and the weighted vector quantization of the continuous pitch coefficients. The techniques allows the faster computation of the optimal combination pitch--continuous coefficient values without substantial loss of optimal results.

Abstract: Method of preparing data, in particular voice signal parameters for transmission at a low bit rate, identical signal parameters are combined interval by interval in quantized form; forfurther bit reduction, bits are suppressed from the total number of bits of at least two intervals, the bit difference to be suppressed being formed on the basis of the total number of unreduced bits with respect to the next-higher power of two; and theprocedure supplies a better voice quality than in the case of changing the number of quantization stages by multiples of 2.

Abstract: The linear predictive analysis method is used in order to determine the spectral parameters representing the spectral envelope of the audiofrequency signal. This method comprises q successive prediction stages, q being an integer greater than 1. At each prediction stage p(1.ltoreq.p.ltoreq.q), parameters are determined representing a predefined number Mp of linear prediction coefficients a.sub.1.sup.p, . . . , a.sub.Mp.sup.p of an input signal of the said stage. The audiofrequency signal to be analysed constitutes the input signal of the first stage.

Abstract: A joint data (features) and channel (bias) estimation framework for robust processing of speech received over a channel is described. A trellis encoded vector quantizer is used as a pre-processor to estimate the channel bias using blind maximum likelihood sequence estimation. Sequential constraint in the feature vector sequence of a speech signal is applied for the selection of the quantized signal constellation and for the decoding process in joint data and channel estimation. A two state trellis encoded vector quantizer is designed for signal bias removal applications.

Abstract: An apparatus and method of quantizing a sequence of input data vectors using delayed decision switched prediction and vector quantization. The method has the following steps of operation: (a) predicting a next vector element from said sequence of input data vectors to generate a set of prediction vectors; (b) subtracting the set of prediction vectors from the next vector element to generate a set of prediction error vectors; (c) multi-stage vector quantizing the set of prediction error vectors to generate a set of quantized prediction error vectors with each of the stages having at least one of the tables and local decision means to generate a final quantization error vector according to a predetermined distance measure; (d) selecting one predictor out of the set of predictors from the switched prediction step and selecting, for each of the stages, at least one entry from the set of tables of the vector quantization step according to the predetermined distance measure, generating a quantized data vector.

Abstract: In a vector quantization apparatus for expressing a target vector by using a code vector designated by an index, an error evaluating section performs error evaluation for a code vector without considering a code error of the index and error evaluation with considering the code error, a first selecting section selects a small number of indexes from a larger number of indexes on the basis of an evaluation result without considering the code error, and a second selecting section selects, on the basis of an evaluation result with considering the code error, an index used to express the target vector from a small number of indexes selected by the first selecting section.

Abstract: An apparatus synchronizes a voice coder and a voice decoder which are of the vector-coding type in order to prevent a false synchronization even when a signal having the same period as a string of synchronizing bits is inputted. A noise component adding unit adds a noise component to an input voice signal. Therefore, even if the input voice signal has the same period as that of a string of synchronizing bits and is completely periodic, the periodicity of the input voice signal is lost by the added noise component. Based on the input voice signal which is no longer periodic, a vector-coding unit, a quantizing signal vector generating unit, and a code book index transmitting unit generate code book indexes and transmit the generated code book indexes to a voice decoder. Therefore, the voice decoder is prevented from developing a false synchronization.

Abstract: A vector quantizer for a speech coder for coding speech signals at low bit rates. The vector quantizer includes an auto-correlation calculation circuit for calculating an impulse response of a weighting function for each sub-interval of an input signal vector. The vector quantizer also includes a weighted cross-correlation calculation circuit for calculating a weighted cross-correlation of the weighted input signal vector and the weighted codevector having a code length equal to that of the input signal vector. The vector quantizer further includes a weighted auto-correlation calculation circuit for calculating an auto-correlation of the weighted codevectors, by using respective auto-correlations of the impulse responses, the codevectors and the cross-correlations.

Abstract: A speech encoding/decoding method calculates a short-term prediction error of an input speech signal that is divided on a time axis into blocks, represents the short-term prediction residue by a synthesized sine wave and a noise and encodes a frequency spectrum of each of the synthesized sine wave and the noise to encode the speech signal. The speech encoding/decoding method decodes the speech signal on a block basis and finds a short-term prediction residue waveform by sine wave synthesis and noise synthesis of the encoded speech signal. The speech encoding/decoding method then synthesizes the time-axis waveform signal based on the short-term prediction residue waveform of the encoded speech signal.

Abstract: The present invention is a speech processing system and method which includes a codebook generator (26) for generating a reference vector quantization codebook which describes a reference environment and for generating at least one secondary vector quantization codebook which describes at least one secondary environment. The secondary vector quantization codebooks are generated using the reference vector quantization codebook. A speech recognizer (38), which is trained using the reference vector quantization codebook, is also included. A pre-processor (32) accepts as input speech collected from an unknown environment and pre-processes that speech, before input to the speech recognizer (38), using the reference vector quantization codebook and an adaptation of the secondary vector quantization codebooks.

Abstract: A method for recognizing speech elements (e.g., phones) in utterances includes the following steps. Based on acoustic frequency, at least two different acoustic representatives are isolated for each of the utterances. From each acoustic representative, tentative decision information on the speech element in the corresponding utterance is derived. A final decision on the speech element in the utterance is then generated, based on the tentative decision information derived from more than one of the acoustic representatives.

Abstract: A speech signal transmitting receiving apparatus, such as a portable telephone set, includes a speech signal transmitting encoding circuit, a noise domain detection unit, a noise level detection unit and a controller. The speech signal transmitting encoding circuit compresses input speech signals by digital signal processing at a high efficiency. The noise domain detection unit detects the noise domain using an analytic pattern produced by the speech signal transmitting encoding circuit. The noise level detection unit detects the noise level of the noise domain detected by the noise domain detection unit. The controller controls the received sound volume responsive to the noise level detected by the noise level detection unit.

Abstract: In method for detecting a speech period in a high-noise environment, the variation in the spectrum of an input signal per unit time is calculated over an analysis frame period, and when the frequency of spectrum variation falls in a predetermined range, the input signal of that frame is decided to be a speech signal.

Abstract: A CELP speech decoder includes a first portion comprising an adaptive codebook and a second portion comprising a fixed codebook. The CS-ACELP decoder generates a speech excitation signal selectively based on output signals from said first and second portions when said decoder fails to receive reliably at least a portion of a current frame of compressed speech information. The decoder does this by classifying the speech signal to be generated as periodic (voiced) or non-periodic (unvoiced) and then generating an excitation signal based on this classification. If the speech signal is classified as periodic, the excitation signal is generated based on the output signal from the first portion and not on the output signal from the second portion. If the speech signal is classified as non-periodic, the excitation signal is generated based on the output signal from said second portion and not on the output signal from said first portion.

Abstract: A speech signal has its characteristics extracted and encoded (16), transmitted over a limited-data-rate path (18) and is decoded (20) and synthesized (22) at the receiving end. The characteristics include line spectral frequencies (LSF), pitch and jitter. The LSF are extracted by autoregression, and splitvector quantized (SVQ) in a single frame, and, in parallel, in blocks of two, three and four frames. The SVQ codes have equal length and are evaluated for distortion in conjunction with a threshold. The threshold is varied in such a manner as tend to select for transmission those codewords which maintain a constant data rate into a transmit buffer. A single-bit jitter bit, and encoded pitch value, are product coded with the selected LSF codeword, and all are transmitted over the data path (18) to the receiver. The receiver decodes the characteristics, and controls a pitch generated (1226) in response to the pitch value and a random pitch jitter in response to the jitter bit.

Abstract: A speech parameter encoder capable of encoding spectrum parameters at a bit rate of 1 kb/s or less with comparatively small amount of operations and memory capacity. A spectrum parameter calculation unit 130 derives a spectrum parameter representing the spectrum envelope of a discrete input speech signal through division thereof into frames each having a predetermined time length. A weighted coefficient calculation unit 150 derives a weighted coefficient corresponding to an auditory masking threshold value through derivation thereof from the speech signal. A spectrum parameter quantization unit 160 receives the weighted coefficient and the spectrum parameter and quantizes the spectrum parameter through search of a codebook such as to minimize the weighting distortion based on the weighted coefficient.

Abstract: A method and apparatus are disclosed for robust, text-independent (and text-dependent) speaker recognition in which identification of a speaker is based on selected spectral information from the speaker's voice. Traditionally, speaker recognition systems (i) render a speech sample in the frequency domain to produce a spectrum, (ii) produce cepstrum coefficients from the spectrum, (iii) produce a codebook from the cepstrum coefficients, and (iv) use the codebook as the feature measure for comparing training speech samples with testing speech samples. The present invention, on the other hand, introduces the important and previously unknown step of truncating the spectrum prior to producing the cepstrum coefficients. Through the use of selected spectra as the feature measure for speaker recognition, the present invention has been shown to yield significant improvements in performance over prior art systems.

Abstract: A speech coding system employing an adaptive codebook model of periodicity is augmented with a pitch-predictive filter (PPF). This PPF has a delay equal to the integer component of the pitch-period and a gain which is adaptive based on a measure of periodicity of the speech signal. In accordance with an embodiment of the present invention, speech processing systems which include a first portion comprising an adaptive codebook and corresponding adaptive codebook amplifier and a second portion comprising a fixed codebook coupled to a pitch filter, are adapted to delay the adaptive codebook gain; determine the pitch filter gain based on the delayed adaptive codebook gain, and amplify samples of a signal in the pitch filter based on said determined pitch filter gain. The adaptive codebook gain is delayed for one subframe. The pitch filter gain equals the delayed. adaptive codebook gain, except when the adaptive codebook gain is either less than 0.2 or greater than 0.8.

Abstract: The present invention concerns efficient quantization of more than one LPC spectral models per frame in order to enhance the accuracy of the time-varying spectrum representation without compromising on the coding-rate. Such efficient representation of LPC spectral models is advantageous to a number of techniques used for digital encoding of speech and/or audio signals.